Policy revisions and concerns
Finance Director Lane Millar said the policy was largely modeled on Sonoma County’s framework, with additional language added by Eureka staff. Training requirements remain broad for now, he said, because available courses are still being developed by the city’s insurance and risk‑management pool.
“The definitions of artificial intelligence, limited memory, AI, generative AI, and agentic AI were updated. The roles and responsibilities section was updated to acknowledge responsibilities regarding the accuracy of AI-assisted work,” Millar said. “Language was added in the limited use section to disallow agentic AI use without prior approval. A section titled Authorization and Training was added to restrict AI use to specific positions, as determined by the department director, and to provide AI training when appropriate.”
Eureka City Mayor Kim Bergal reminded the council, “I just did want to clarify that, this particular topic tonight is about employee use of AI, not whether we agree with it or disagree with it or what we personally think about it.”
Councilmember Kati Moulton expressed concern with the policy and asked where the revised policy mentions that employees should not use AI tools to complete work that is out of their scope or expertise. She also noted the training portion has not been defined.
“I think this is significant, specifically because AI is capable of writing computer code. For example, if you can’t check that code to make sure that it is not going to cause some harmful effect that you’re unaware of, then you shouldn’t be doing that. We wouldn’t have people putting together code that didn’t know how to do it,” said Moulton.
Image‑use restrictions
Council members spent significant time debating how far to go in limiting image or video uploads to AI tools. Moulton pushed for a near-total ban on uploading photos, videos or artwork into AI systems, citing copyright, privacy, ethical and environmental concerns.
“I don’t want to throw city clerk Powell under the bus, but we were looking at something to put onto social media, and there was a picture that we had in our stock photos of Carson, and, uh, utilized AI to add a camera in the picture to make it seem like a movie scene type of thing,” City Manager Miles Slattery said. “Full disclosure, that has been done. It was done for our film permit web page.”
Others favored a narrower approach. Councilmember Scott Bauer said he doesn’t have issues with using AI to edit a photo’s appearance if you are not adding anything generated into it.
Councilmember Leslie Castellano shared a different concern with the council.
“I don’t mind if a photograph of a building is put into AI,” said councilmember Castellano. “But if we’ve hired a photographer, that should be part of the consent form, whether the photo can be put in AI. And generally speaking, I don’t think we should put, like, constituents and residents’ pictures into AI without their consent.”
Environmental and economic stakes
Moulton compared the environmental footprint of AI systems to early automobile manufacturing and warned that the technology’s rapid adoption could reshape local jobs.
“I really appreciate that we are having this conversation. AI is an incredibly powerful set of tools, and when something is so readily accessible and so easy to use, it’s easy to lose sight of what a big deal it can be,” said Moulton. “AI is economically and environmentally on par with the manufacturing of gas automobiles. We are making decisions right now that are going to impact our environment for centuries.”
A public commenter Zack, identifying himself as a computer programmer, urged the council to adapt tougher regulations on employees generating AI images or videos at work.
“I think it would be really smart to outright block city employees from generating AI videos or images entirely,” he said. “Video and image systems are built as entertainment products, and they’re not very good at productivity. They’re both the most environmentally disruptive and the most costly to use, and they’re huge liabilities.”
Moulton read aloud a statement from the IT department that addressed the city policy on usage of generative AI, along with concerns about the growing AI technologies.
“It does not document the department’s concern with substantially increasing costs of computer equipment due to the AI bubble…Nor does it document the department’s concern with potential CGIS, HIPA, PCI violations…nor does it document any employees’ personal concerns with the industry in general environmental concerns, human right concerns and other economic concerns,” she said.
Final amendments and vote
After a lengthy deliberation, the council agreed on the issues with the previous policy and agreed to a resolution. Councilmember G. Mario Fernandez, attending via Zoom, moved to adopt the rewritten policy.
A substitute motion followed to include two additional amendments:
- prohibiting employees from using AI to perform work outside their job expertise, and
- requiring written consent before uploading images of people or artists created materials.
The vote came moments before a twist. Soon after the council’s extended discussion on limiting the usage of AI tools in the workplace, city employee Jacob Rosen, who was presenting on the CARE program, showed a slideshow that included a disclaimer on AI usage and contained several AI-generated images.
Maranda Vargas can be reached at (707) 441-0504
© 2026 Times-Standard, Eureka, Calif.. Visit www.times-standard.com. Distributed by Tribune Content Agency, LLC.
